Back 
The advent of cloud applications led to a new generation of phishing attacks (named OAuth phishing or consent phishing) where, rather than stealing the user credentials, threat actors aim to obtain an authorization token via a rogue cloud app that allows them to perform harmful activities on the victim’s cloud environment. These activities include: reading emails, sending emails on behalf of the victim itself, reading the calendar, changing profile and email settings, and many others permissions that can be exploited to steal confidential information from the target.
The so-called user experience is similar to a traditional phishing attack. The victim receives a malicious link, for example via a spear phishing email, but instead of being presented with a form mimicking a legitimate app eager to steal their credentials, they are presented with an authentication page where a seemingly legitimate cloud app asks to access their cloud environment with specific permissions on behalf of the user, permissions that the victims do not double check and end up delivering attackers the keys of the kingdom Multi-factor authentication does not help to mitigate this attack, since the user must authenticate to the target app to provide consent to the malicious application. Then once the authorization token is issued, the only way to remediate is to delete the malicious app from the target environment or, in other terms, revoke the token.